data augmentation is one of the most important tools in training modern deep neural networks. Recently, great advances have been made in searching for optimal augmentation policies in the →
本文介绍了使用Online Hyper-parameter Learning for Auto-Augmentation解决基于离线搜索增广策略所存在的高昂成本问题。在CIFAR-10和ImageNet数据集上,我们的方法取得了显著的准确率提高,并且速度分别比现有方法快了60倍和24倍,同时保持了有竞争力的准确性。